All sessions of Transformation 2021 are now available on demand. Now look.

One of NASA’s biggest challenges is identifying where data science skills reside in the organization. Data science is just a new discipline – it’s also a fast-growing one. Knowledge for each role is constantly changing due to technical and business demands.

That’s where David Meza, head of the executive branch of People Analytics and a senior NASA data scientist, believes graph technology can help. His team is creating a Talent Mapping Database using Neo4J technology to create a knowledge graph to show the relationship between people, skills and projects.

Meza and his team are currently working on the implementation phase of the project. They finally plan to formalize the end-user app and create an interface to help NASA people find talent and job opportunities. Meja told VentureBit more about the project.

VentureBet: What is the broader purpose of this data lead project?

David Meza: It’s a look at how we can identify the skills, knowledge and abilities, tasks and techniques within a business or work role. How can we translate it into an employee? How do we combine it with their training? And how do we connect it with projects and programs? All that work is a matter of a relationship that can be connected by certain elements that connect it all together – and that’s where the graph comes from.

VentureBet: Why did you decide to go with Neo 4J rather than internal development?

Meza: I think there really was nothing out there that provided what we were looking for, so that’s part of it. The other part of the process is that we have the specific information we are looking for. It’s not very common. And so we needed to create something that is more necessary for our concepts, our ideas and the very specific things we do at NASA around spaceflight, operations and the like.

VentureBet: What is the timeline for the Neo 4J release?

Meza: We are still in the implementation phase. The first six to eight months were spent researching and developing and ensuring we had proper access to data. Like any other project, it is our most difficult task – to make sure we have the right access, to have the right information and to think about how everything is related. When we looked at that, we also worked in parallel on other issues: what would the model look like, what algorithms would we use, and how would we train these models? We’ve now got the data in the graph system and we’re starting to generate a beta phase of an application. This summer, by the end of the year, we’re in the process of making that application izing formal to make it more of an end-user interface.

VentureBit: What is the technical process behind the implementation of Neo 4?

Meza: The first part was trying to think about what our business classification would be. We looked at: “How do we recognize business? What is the DNA of a business? “And in the same way, we looked at it from an employee perspective, from a training perspective, and from a program or project perspective. So in simple terms, we broke everything down into three different categories for each business: part of knowledge, one skill, and work.”

VentureBet: How are you using those categories to create data models?

Meza: If you can start identifying people who have great knowledge in the process of natural language, for example, and the skills they need to do a job, you can say from a business point of view that specialized workers need special skills and abilities. Fortunately, there is a database called O * Net from the Labor Department, which contains details of hundreds of businesses and their elements. Those elements include knowledge, skills, abilities, functions, employee characteristics, licensing and education. So that was the basis for our Neo4j graph database. Then we did the same with training. Within the training, you are learning a part of the knowledge; To learn that part of knowledge, you will gain a skill; And to gain that skill, you are going to do exercises or tasks to master that skill. And it’s the same for programs: we can connect back to what knowledge, skills and tasks the person needs for each project.

Venturebet: How will you train the model over time?

Meza: We are beginning to look at NASA-specific capabilities and task roles to assign to employees. Our next step is to validate and test employees – around knowledge, skills, abilities, functions and technologies – whether what we imagine based on the model is either true or false. After that, we will use that feedback to train the model so that it can do a little better. That’s what we hope to do in the next few months.

Venturebeat: What would this approach mean for identifying talent at NASA?

Meza: I think it will give employees a chance to see what’s out there that will interest them in advancing their careers. If they want to make a career change, for example, they can see where they are in the process. But I also think it will help us better organize our people in our organization, and we can help track and predict where we may lose skills, where we need to migrate our programs and our migration. Skills need to change depending on the basis. Mission due to change of administration. So I think it will make us a little more agile and make it easier to move our workforce.

VentureBet: Do you have any best practice lessons for implementing Neo4j?

Meza: I think the biggest lesson I have learned so far is to identify the many data sources that can help you provide some information. Start small – you don’t need to know everything right away. When I look at knowledge graphs and graph databases, the beauty is that you can add and remove information more easily than in a relational database system, where you should know the schema upfront. Within a graph database or knowledge graph, you can easily retrieve information without messing up your schema or your data model. Adding more information only enhances your model. So start small, but think big in terms of what you are trying to do. See how you can develop relationships, and also try to identify latent relationships in your graph based on the information you have about those data sources.


VentureBet’s mission is to become a digital town square for technical decision makers to gain knowledge about changing technology and transactions. Our site provides essential information on data technology and strategies to guide you as you lead your organizations. We invite you to become a member of our community for:

  • Up-to-date information on topics of interest to you
  • Our newsletters
  • Gated idea-leader content and discounted access to our precious events, e.g. Transformation 2021: Learn more
  • Networking features and more

Become a member